| Label-Retrieval-Augmented Diffusion Models for Learning from Noisy Labels | ✓ Link | 84.16 | | 82.56 | | LRA-diffusion (CLIP ViT) | 2023-05-31 |
| Two Wrongs Don't Make a Right: Combating Confirmation Bias in Learning with Label Noise | | 81.84 | 94.12 | 75.48 | 93.76 | Robust LR | 2021-12-06 |
| Sample Prior Guided Robust Model Learning to Suppress Noisy Labels | ✓ Link | 81.47 | 94.03 | 75.45 | 93.11 | PGDF (Inception-ResNet-v2) | 2021-12-02 |
| SSR: An Efficient and Robust Framework for Learning with Unknown Label Noise | ✓ Link | 80.92 | 92.80 | 75.76 | 91.76 | SSR | 2021-11-22 |
| Bootstrapping the Relationship Between Images and Their Clean and Noisy Labels | ✓ Link | 80.88 | 92.76 | 75.96 | 92.20 | BtR | 2022-10-17 |
| CoDiM: Learning with Noisy Labels via Contrastive Semi-Supervised Learning | | 80.88 | 92.48 | 76.52 | 91.96 | CoDiM-Sup (Inception-ResNet-v2) | 2021-11-23 |
| Learning with Neighbor Consistency for Noisy Labels | ✓ Link | 80.5 | | | | NCR+Mixup+DA (ResNet-50) | 2022-02-04 |
| CMW-Net: Learning a Class-Aware Sample Weighting Mapping for Robust Deep Learning | ✓ Link | 80.44 | 93.36 | 77.36 | 93.48 | CMW-Net-SL+C2D | 2022-02-11 |
| Dynamic Loss For Robust Learning | ✓ Link | 80.12 | 93.64 | 74.76 | 93.08 | Dynamic Loss (Inception-ResNet-v2) | 2022-11-22 |
| CoDiM: Learning with Noisy Labels via Contrastive Semi-Supervised Learning | | 80.12 | 93.52 | 77.24 | 92.48 | CoDiM-Self (Inception-ResNet-v2) | 2021-11-23 |
| Selective-Supervised Contrastive Learning with Noisy Labels | ✓ Link | 79.96 | 92.64 | 76.84 | 93.04 | Sel-CL+ (ResNet-18) | 2022-03-08 |
| Class Prototype-based Cleaner for Label Noise Learning | ✓ Link | 79.63±0.08 | 93.46±0.10 | 75.75±0.14 | 93.49±0.25 | CPC | 2022-12-21 |
| PSSCL: A progressive sample selection framework with contrastive loss designed for noisy labels | ✓ Link | 79.56 | 94.84 | 79.68 | 95.16 | PSSCL (130 epochs) | 2024-12-18 |
| Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy Labels | ✓ Link | 79.42 ± 0.34 | 92.32 ± 0.33 | 78.57 ± 0.37 | 93.04 ± 0.10 | DivideMix with C2D (ResNet-50) | 2021-03-25 |
| Faster Meta Update Strategy for Noise-Robust Deep Learning | ✓ Link | 79.4 | 92.80 | 77 | 92.76 | FaMUS | 2021-04-30 |
| Learning with Neighbor Consistency for Noisy Labels | ✓ Link | 79.4 | | | | NCR+Mixup (ResNet-50) | 2022-02-04 |
| Centrality and Consistency: Two-Stage Clean Samples Identification for Learning with Instance-Dependent Noisy Labels | ✓ Link | 79.36 | 93.64 | 76.08 | 93.86 | CC | 2022-07-29 |
| Generalized Jensen-Shannon Divergence Loss for Learning with Noisy Labels | ✓ Link | 79.28 | 91.22 | 75.50 | 91.27 | GJS (ResNet-50) | 2021-05-10 |
| NGC: A Unified Framework for Learning with Open-World Noisy Data | | 79.16 | 91.84 | 74.44 | 91.04 | NGC (Inception-ResNet-v2) | 2021-08-25 |
| Twin Contrastive Learning with Noisy Labels | ✓ Link | 79.1 | 92.3 | 75.4 | 92.4 | TCL | 2023-03-13 |
| LongReMix: Robust Learning with High Confidence Samples in a Noisy Label Environment | ✓ Link | 78.92 | 92.32 | | | LongReMix (Inception-ResNet-v2) | 2021-03-06 |
| Multi-Objective Interpolation Training for Robustness to Label Noise | ✓ Link | 78.76 | | | | MOIT+ (ResNet-18) | 2020-12-08 |
| PSSCL: A progressive sample selection framework with contrastive loss designed for noisy labels | ✓ Link | 78.52 | 93.80 | 79.40 | 94.84 | PSSCL (120 epochs) | 2024-12-18 |
| CMW-Net: Learning a Class-Aware Sample Weighting Mapping for Robust Deep Learning | ✓ Link | 78.08 | 92.96 | 75.72 | 92.52 | CMW-Net-SL | 2022-02-11 |
| Early-Learning Regularization Prevents Memorization of Noisy Labels | ✓ Link | 77.78 | 91.68 | 70.29 | 89.76 | ELR+ (Inception-ResNet-v2) | 2020-06-30 |
| ScanMix: Learning from Severe Label Noise via Semantic Clustering and Semi-Supervised Learning | ✓ Link | 77.72 | | | | ScanMix (Inception-ResNet-v2) | 2021-03-21 |
| Robust Long-Tailed Learning under Label Noise | | 77.64 | 92.44 | 74.64 | 92.48 | ROLT+ (Inception-ResNet-v2) | 2021-08-26 |
| Sample Selection with Uncertainty of Losses for Learning with Noisy Labels | | 77.53 | | | | CNLCU-S + DivideMix (Inception-ResNet-v2) | 2021-06-01 |
| Hard Sample Aware Noise Robust Learning for Histopathology Image Classification | ✓ Link | 77.52 | | | | HSA-NRL(Inception-ResNet-v2) | 2021-12-05 |
| Confidence Adaptive Regularization for Deep Learning with Noisy Labels | | 77.41 | 92.25 | 74.09 | 92.09 | CAR | 2021-08-18 |
| DivideMix: Learning with Noisy Labels as Semi-supervised Learning | ✓ Link | 77.32 | 91.64 | 75.20 | 91.64 | DivideMix (Inception-ResNet-v2) | 2020-02-18 |
| Learning with Neighbor Consistency for Noisy Labels | ✓ Link | 77.1 | | | | NCR (ResNet-50) | 2022-02-04 |
| DivideMix: Learning with Noisy Labels as Semi-supervised Learning | ✓ Link | 76.32 ±0.36 | 90.65 ±0.16 | 74.42 ±0.29 | 91.21 ±0.12 | DivideMix (ResNet-50) | 2020-02-18 |
| DivideMix: Learning with Noisy Labels as Semi-supervised Learning | ✓ Link | 76.08 | | | | DivideMix (ResNet-18) | 2020-02-18 |
| Beyond Synthetic Noise: Deep Learning on Controlled Noisy Labels | ✓ Link | 76.0 | 90.2 | 72.9 | 91.1 | MentorMix (Inception-ResNet-v2) | 2019-11-21 |
| Noisy Concurrent Training for Efficient Learning under Label Noise | ✓ Link | 75.16 | 90.77 | 71.73 | 91.61 | NCT (Inception-ResNet-v2) | 2020-09-17 |
| Robust and On-the-fly Dataset Denoising for Image Classification | | 74.6 | 90.6 | 66.7 | 86.3 | ODD (Inception-ResNet-v2) | 2020-03-24 |
| Coresets for Robust Training of Neural Networks against Noisy Labels | | 72.40 | 89.56 | 67.36 | 87.84 | Crust (Inception-ResNet-v2) | 2020-11-15 |
| Understanding and Utilizing Deep Neural Networks Trained with Noisy Labels | ✓ Link | 65.2 | 85.3 | 61.6 | 85.0 | Iterative-CV (Inception-ResNet-v2) | 2019-05-13 |
| Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels | ✓ Link | 63.58 | 85.20 | 61.48 | 84.70 | Co-teaching (Inception-ResNet-v2) | 2018-04-18 |
| Dimensionality-Driven Learning with Noisy Labels | ✓ Link | 62.68 | 84.00 | 57.80 | 81.36 | D2L (Inception-ResNet-v2) | 2018-06-07 |
| Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach | ✓ Link | 61.12 | 82.68 | 57.36 | 82.36 | F-Correction (Inception-ResNet-v2) | 2016-09-13 |
| Robust Temporal Ensembling for Learning with Noisy Labels | | | | 80.84 | 97.24 | RTE (Inception-ResNet-v2) | 2021-09-29 |
| MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels | ✓ Link | | | 63.8 | 85.8 | MentorNet (Inception-ResNet-v2) | 2017-12-14 |
| Normalized Loss Functions for Deep Learning with Noisy Labels | ✓ Link | | | 62.64 | | NCE+RCE (ResNet-50) | 2020-06-24 |
| Normalized Loss Functions for Deep Learning with Noisy Labels | ✓ Link | | | 62.36 | | NCE+MAE (ResNet-50) | 2020-06-24 |
| Robust early-learning: Hindering the memorization of noisy labels | | | | 61.85 | | CDR (Inception-ResNet-v2) | 2021-01-01 |